Quantifying Volatility in VaR Models

VaR is a method that is widely used to analyze the risk associated with any portfolio. The risk analysis is very much important in the modern day investment. It is also defined as a technique used to determine the probability of loss that the asset may incur by using the historical data of the asset performance. This VaR technique is widely used by banks, security investors and investors involved in commodity trading. VaR is mainly used in trading and hedging decision making. It is used widely as a technical analysis in security trading. For example if a asset has a VaR value of 5 % of 1 million it assumed that the asset will make a million loss once in every 20 time. The five primary uses of VaR are financial risk management, risk assessment, financial control and financial reporting. VaR is at times used in non financial places as well.

Quantifying volatility in VaR model consist of two types. One is primarily used for risk measurement and the other one is primarily used for risk management. The manager uses the VaR as a system. They compare the published value with the statistical value and determine the risk associated with the asset. The main problem with this being the input error misleads the outcome often. This is called as VaR error or VaR break down. This break down happens occasionally. It is difficult to predict this break down. The main reason for the inability to predict this break down is due to unconditional occurrence of it. It will occur at any time which cannot be predicted. VaR break down is often termed as independent event.

VaR predicts that the occurrence of loss is very less when the volatility is low and it is very high when the volatility is very high, thus volatility in the market can be used to determine the occurrence of loss using this method. VaR always take values in the range of 5% to 1 % and it has a time period of two week horizon. For higher risk asset the % might slightly go up.

Question

• What is VaR?
• What are the various uses of VaR?